How to migrate from Amazon Nova 1 to Nova 2 in Bedrock: model mapping and new capabilities

AWS has published a detailed guide for migrating from Amazon Nova 1 to Nova 2 within the Amazon Bedrock platform.

What

Changed in Nova 2 Nova 2 is the next generation of AWS models with improved quality, speed, and inference cost metrics. The transition affects several levels: model identifiers, API parameters, and configuration of new features. Key changes during migration: Updated model ID mapping for each version of the Nova family Changes in request structure via Converse API New parameters for managing Nova 2 capabilities Support for expanded context window * Updated token limits and throughput ## How the Transition Looks AWS recommends using the Converse API as the primary migration path — it abstracts differences between model versions and simplifies code updates.

The guide provides concrete examples: how to replace the model ID, how to adapt call parameters, and how to enable features that were not available in Nova 1. For teams using streaming generation or multimodal input data, AWS separately describes changes in image and document processing.

Migration

Planning Checklist At the end of the guide is a structured migration checklist: from auditing current integrations to testing in a staging environment before moving production traffic. AWS emphasizes that most existing Bedrock applications require minimal code changes.

What

This Means Nova 2 offers the best price-to-quality ratio for enterprise tasks on AWS, and migrating to it is a planned operation, not a complex migration: updating a few lines of config and model ID solves most of the work.